I have a very novice question: should we compare between graphs, or should we consider each an independent test (because the setup or conditions might have changed between tests)? Since these are actual measurements (not theoretical calculations), I would assume they're very sensitive to...